linux-kernel - Re: [Intel-gfx] [PATCH] drm/i915: zero fill vma name buffer

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 16 Sep 2021 09:23:20 +0100
From:   Tvrtko Ursulin <tvrtko.ursulin@...ux.intel.com>
To:     Tim Gardner <tim.gardner@...onical.com>,
        linux-kernel@...r.kernel.org
Cc:     Jani Nikula <jani.nikula@...ux.intel.com>,
        Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>,
        Rodrigo Vivi <rodrigo.vivi@...el.com>,
        David Airlie <airlied@...ux.ie>,
        Daniel Vetter <daniel@...ll.ch>,
        intel-gfx@...ts.freedesktop.org, dri-devel@...ts.freedesktop.org
Subject: Re: [Intel-gfx] [PATCH] drm/i915: zero fill vma name buffer


On 15/09/2021 20:23, Tim Gardner wrote:
> In capture_vma() Coverity complains of a possible buffer overrun. Even
> though this is a static function where all call sites can be checked,
> limiting the copy length could save some future grief.
> 
> CID 93300 (#1 of 1): Copy into fixed size buffer (STRING_OVERFLOW)
> 4. fixed_size_dest: You might overrun the 16-character fixed-size string c->name
>     by copying name without checking the length.
> 5. parameter_as_source: Note: This defect has an elevated risk because the
>     source argument is a parameter of the current function.
> 1326        strcpy(c->name, name);
> 
> Fix any possible overflows by using strncpy(). Zero fill the name buffer to
> guarantee ASCII string NULL termination.
> 
> Cc: Jani Nikula <jani.nikula@...ux.intel.com>
> Cc: Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>
> Cc: Rodrigo Vivi <rodrigo.vivi@...el.com>
> Cc: David Airlie <airlied@...ux.ie>
> Cc: Daniel Vetter <daniel@...ll.ch>
> Cc: intel-gfx@...ts.freedesktop.org
> Cc: dri-devel@...ts.freedesktop.org
> Cc: linux-kernel@...r.kernel.org
> Signed-off-by: Tim Gardner <tim.gardner@...onical.com>
> ---
>   drivers/gpu/drm/i915/i915_gpu_error.c | 7 ++++---
>   1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
> index 9cf6ac575de1..154df174e2d7 100644
> --- a/drivers/gpu/drm/i915/i915_gpu_error.c
> +++ b/drivers/gpu/drm/i915/i915_gpu_error.c
> @@ -1297,10 +1297,11 @@ static bool record_context(struct i915_gem_context_coredump *e,
>   	return simulated;
>   }
>   
> +#define VMA_NAME_LEN 16
>   struct intel_engine_capture_vma {
>   	struct intel_engine_capture_vma *next;
>   	struct i915_vma *vma;
> -	char name[16];
> +	char name[VMA_NAME_LEN];
>   };
>   
>   static struct intel_engine_capture_vma *
> @@ -1314,7 +1315,7 @@ capture_vma(struct intel_engine_capture_vma *next,
>   	if (!vma)
>   		return next;
>   
> -	c = kmalloc(sizeof(*c), gfp);
> +	c = kzalloc(sizeof(*c), gfp);
>   	if (!c)
>   		return next;
>   
> @@ -1323,7 +1324,7 @@ capture_vma(struct intel_engine_capture_vma *next,
>   		return next;
>   	}
>   
> -	strcpy(c->name, name);
> +	strncpy(c->name, name, VMA_NAME_LEN-1);

GCC is supposed to catch any problems here as you say in the commit message.

But to fix I suggest a single line change to strlcpy(c->name, name, 
sizeof(c->name)) which always null terminates as bonus.

Probably same in i915_vma_coredump_create() which with strncpy would 
have a theoretical chance of attempting to copy over a 
non-null-terminated string.

Regards,

Tvrtko

>   	c->vma = vma; /* reference held while active */
>   
>   	c->next = next;
>